Increasing the efficiency of bacterial transcription simulations: When to exclude the genome without loss of accuracy